Memory modules are tested before they are mounted in PCs or sold. Such testing is preferably done in an automated manner and in serial production, with a number of modules being introduced simultaneously into a testing apparatus, to be subsequently contacted one by one with electrical testing terminals for performing a functional test. Depending on the result of the test, the modules are subsequently sorted into properly working and improperly working modules. U.S. Pat. No. 5,704,489 discloses an apparatus for testing memory modules. This apparatus comprises a magazine into which a plurality of modules stacked onto each other can be introduced. This magazine stands substantially vertically on the apparatus, so that the stack of memory modules rests on the underside of the magazine. At the underside of the magazine is a mechanism for dropping the modules one by one from the magazine onto a table. On the table is a slide which proceeds to push the modules into the test terminals.
The apparatus of U.S. Pat. No. 5,704,489 is of modular design, so that it is suitable to be converted if after handling a first batch of memory modules of a first type, a second batch of memory modules of a second type having different dimensions or a different connection are to be tested.
The apparatus mentioned has a number of disadvantages. Firstly, the number of memory modules that can be simultaneous placed in the magazine is rather limited, in particular if modern types of memory modules are involved. This is because these modern types of modules comprise, on the carrier surface thereof, adjacent the edge of the carrier surface, a component (an EPROM) which projects higher above the carrier surface than the other components. As a consequence, the memory modules fan out when being stacked: the stack becomes highest on the side of the projecting component and much lower on the opposite side. If significant fanning is involved, that is, fanning through a large angle and/or cumulative fanning of more than a few a circuit modules, this limits the maximum number of memory modules that can be properly stacked. Further, the weight of high stacks presses onto the lower modules, which are supported only on the sides and therefore may easily be damaged. This also limits the number of modules that can be stacked.
Secondly, the known machine has the disadvantage that the modules, by falling onto the table, can get damaged. Thirdly, the resetting in between batches of different types, despite the modular design, is still considerable, because the height of the table must be adjusted to the height of the components on the carrier surface. Fourthly, the mechanism of the known machine does not allow modules that differ mutually in the height of the components on them, to be jointly placed in the magazine for testing. Fifthly, the mechanism of the known machine only lends itself for testing one module at a time.